Phosphatidylethanolamine-binding protein 1 (PEBP-1) was first isolated as a cytosolic, 23 kDa protein from bovine brain and was also found in soluble extracts from human platelets and rodent brains (Bernier and Jolles 1984). A series of studies over the next decade led to its identification as the precursor for hippocampal cholinergic neurostimulating peptide (HCNP), an acetylated eleven amino acid long peptide conserved in the rat, human, and bovine genome (Tohdoh et al. 1995). Although, it was originally named for its binding affinity for the phospholipid phosphatidylethanolamine, a breakthrough report characterized PEBP-1 as a Raf-1 kinase binding protein with the ability to inhibit the mitogen-activated protein kinase (MAPK) pathway; hence the name Raf kinase inhibitory protein (RKIP) (Yeung et al. 1999). These seminal studies catalyzed a growth of interest for PEBP-1/RKIP, and since then PEBP-1/RKIP has been implicated in a myriad of cellular and physiological processes by way of its privileged position in various signaling pathways.
Structure and Expression
In humans, pebp-1 is located on chromosome 12q24.23. Four exons encode the full-length, 187 amino acid long PEBP-1/RKIP, with the first exon encoding for HCNP. The sequence of HCNP is acetyl-Ala-Ala-Asp-Ile-Ser-Gln-Trp-Ala-Gly-Pro-Leu. Its release in the hippocampus was found to be dependent on NMDA receptor activation (Ojika et al. 2000). The promoter of pebp-1/rkip lacks a TATA box and instead uses CpG islands, a common feature of housekeeping genes (Okita et al. 2009). There are also enhancer elements within the promoter from positions −596 to −417 relative to the TSS. In mammals and under basal conditions, PEBP-1/RKIP is expressed in virtually all tissues of the body. However, the highest levels are present in the testis, brain, and adrenal gland (Frayne et al. 1999; Uhlén et al. 2015). Levels of PEBP-1/RKIP in the brain are considerably lower during embryonic development compared to adulthood, and decline again at old age (Frayne et al. 1999). Using a mouse strain carrying a β-geo reporter, Theroux and colleagues characterized PEBP-1/RKIP expression in the adult brain and found the highest levels in limbic structures (hypothalamus, hippocampus, nucleus accumbens, amygdala) as well as cerebellar Purkinje cells and olfactory neurons (Theroux et al. 2007). PEBP-1/RKIP knock out mice display a progressive olfaction deficit starting at four months of age (Theroux et al. 2007).
Although structurally compact, the PEBP-1/RKIP protein has some outstanding properties that likely contribute to its functional diversity. PEBP-1/RKIP contains a ligand-binding pocket that is similar to conserved phospholipid-interacting domains present in other PEBP family members. PEBP-1/RKIP is capable of binding to phospholipids in in vitro assays, but whether or not this interaction occurs in vivo remains unclear (Skinner and Rosner 2014). The pocket loop has been shown to bind nucleotides, and there is significant evidence that this subregion as well as neighboring residues mediate binding to signaling molecules such as Raf-1 (Tavel et al. 2012). Five phosphorylation sites on PEBP-1/RKIP have been identified thus far: threonine-42, serine-52, serine-54, serine-99, and serine-153. The most highly characterized phosphorylation at ser-153 has been shown to trigger further structural changes leading to PEBP-1/RKIP homodimerization, which in turn controls its binding specificity for other protein partners (Deiss et al. 2012). However, we are just beginning to understand the functional contributions of these modifications.
Roles in Cell Signaling
Roles in Cell Function
PEBP-1/RKIP plays a role in many cellular functions that are regulated by the MAPK signaling pathway. One of these is cellular senescence, a response to cell damage that is mediated by the tumor suppressor protein p53 and is accompanied by downregulation of MAPK activity. Stress due to DNA damage increases expression of p53. Phosphorylation of p53 at ser-46 triggers its transcriptional activation of the pebp-1/rkip gene (Lee et al. 2013). Furthermore, irradiated senescent tumor cells have been shown to secrete PEBP-1/RKIP, which promotes the migration of neighboring cells (Han et al. 2012). The role of PEBP-1/RKIP in cell motility and migration has been studied in different contexts, beginning with a report where the interaction between Raf-1 and PEBP-1/RKIP was shown to be disrupted by locostatin, a cell migration inhibitor (Zhu et al. 2005). In vitro silencing of PEBP-1/RKIP expression also slowed the rate of epithelial cell migration, whereas overexpressing PEBP-1/RKIP dramatically increased cell motility, conferring a fibroblast-like phenotype (Zhu et al. 2005). Other studies suggest that PEBP-1/RKIP may have a suppressive rather than facilitatory effect on cell migration. For instance, in various types of malignant tumors, expression levels of this gene are inversely correlated with metastasis (Yesilkanal and Rosner 2014). Some studies suggest that PEBP-1/RKIP may suppress cell migration and invasion through signaling networks downstream of the MAPK cascade that may involve chromatin remodeling factors and micro-RNAs (Yesilkanal and Rosner 2014). In another in vitro model using human embryonic cell lines, PEBP-1/RKIP was suggested to suppress cell motility by inhibiting the expression or stability of p21-activated protein kinase 1 (PAK1), β-catenin, and vimentin (Al-Mulla et al. 2011b). In the same study, a role for PEBP1/RKIP in cellular proliferation was investigated. Depleting cells of pebp1/rkip expression lead to an increase in cell division, by overriding mitotic checkpoints and accelerating progression through G1/S and G2/M (Al-Mulla et al. 2011b). PEBP-1/RKIP also plays a role in determining cell fate in neural lineages, as observed in vitro using neuroblastoma cell lines (Hellmann et al. 2010). Although the mechanisms are yet to be elucidated, silencing PEBP-1/RKIP promotes a glial phenotype, whereas its overexpression tips the balance toward a neuronal fate (Hellmann et al. 2010).
In various tissue and cell types, PEBP-1/RKIP expression contributes to their normal functioning, but it may also have deleterious effects under pathological conditions. For example, by regulating β-adrenergic signaling at the level of β1 and β2 receptors in the heart, PEBP-1/RKIP improves cardiac contractility and has antiapoptotic and antifibrotic effects (Schmid et al. 2015). In contrast, PEBP-1/RKIP expression has a negative effect in a mouse model for systemic inflammatory response syndrome (SIRS), a common cause for complications in hospitalized patients. SIRS develops due to an exaggerated release of proinflammatory cytokines. PEBP-1/RKIP reportedly exacerbates this cytokine storm by driving the overproduction of one such cytokine, interferon gamma (IFN-γ), by CD8+ T cells (Wright and Vella 2013).
PEBP-1/RKIP and its derivatives have also been found to play roles in brain function and pathologies. The HCNP peptide stimulates acetylcholine synthesis in the medial septal nucleus by increasing the expression of the enzyme choline acetyltransferase (ChAT) . Overexpressing the precursor protein in murine forebrain neurons leads to a depressive-like phenotype specifically at old age, while overall normal cognitive function is maintained (Matsukawa et al. 2010). Long-term depression (LTD) is another instance where the modulation of the MAPK pathway by PEBP-1/RKIP shapes a physiological function. LTD can be induced in vivo and ex vivo in specific neural circuits, among them the input from the parallel and climbing fibers, to Purkinje cells in the cerebellum. LTD is an electrophysiological phenomenon where short patterns of neural activity have a long-lasting inhibitory effect. Such plasticity requires intracellular calcium release, PKC, and persistent MAPK activity. A study by Yamamoto and collaborators found that phosphorylation of PEBP-1/RKIP by PKC is required to disrupt its interaction with Raf-1 or MEK, promoting persistent MAPK activity (Yamamoto et al. 2012). More recently, a role in long-term potentiation (LTP) was also reported. LTP is the long-lasting facilitation in a neural circuit following a short pattern of activity. LTP in the Schaffer collateral of the hippocampus was investigated by Ohi and collaborators in a murine model. They found that overexpression of PEBP-1/RKIP led to enhancement of LTP, possibly through the action of acetylcholine muscarinic receptors at glutamatergic postsynapses (Ohi et al. 2015).
In mammals, the clock cells of the suprachiasmatic nucleus (SCN), the hypothalamic region responsible for maintaining circadian (∼24 h) rhythmicity, rely on MAPK signaling to synchronize with the environment. In these cells, synchronized and self-sustained daily rhythms in transcription and translation of clock genes such as Period1/2 can be reset by external time cues. Light-induced Period transcription depends on the rapid and transient activation of MAPK/ERK in SCN neurons, leading to a shift in the clock phase as the SCN realigns to match environmental time. In the absence of PEBP-1/RKIP, light-induced MAPK activity in the SCN is prolonged, and the phase-shifting response to light stimuli is consequently exaggerated (Antoun et al. 2012). However, the role of PEBP-1/RKIP in circadian rhythms is yet to be examined in relation to other signaling regulators such as GRKs.
A potential involvement of PEBP-1/RKIP in Alzheimer’s disease (AD) has also been suggested. When investigated in postmortem brains of AD patients, the protein was found to be expressed in Hirano bodies, a neuropathological component of the disease (Ojika et al. 2000).
The role of PEBP-1/RKIP in cancer has been extensively investigated. Many clinical studies have reported downregulation in PEBP-1/RKIP expression in solid tumors of the prostate, breast, colon/rectum, esophagus, stomach, ovary, cervix, liver, kidney, lung, nasopharynx, and brain, as well as cancers of the bone marrow and blood (Lamiman et al. 2014). The relationship between RKIP and tumorigenesis was first studied in prostate cancer, where low PEBP-1/RKIP expression correlated with cancer metastasis and a higher rate of cancer recurrence (Fu et al. 2006). Similar results have been found in other tumor types. Clinical results suggest that PEBP-1/RKIP may be a useful prognostic marker for metastasis, overall survival, and disease-free survival (Lamiman et al. 2014). However, this cannot be generalized to all cancers, as loss of PEBP-1/RKIP appears to be a favorable prognostic indicator for patients with acute myeloid leukemia (Zebisch et al. 2012). The mechanism for loss of PEBP-1/RKIP expression in human cancers is unclear, although gene silencing through promoter methylation is believed to be a contributing factor in some cancer types (Li et al. 2014). Collectively, these findings have cemented PEBP-1/RKIP’s reputation as a metastasis supressor, but also as a potential biomarker for disease prognosis (Lamiman et al. 2014).
First identified as the precursor for HCNP, PEBP-1/RKIP is an evolutionarily conserved, small, and compact protein. Its structure displays a versatile binding pocket that allows it to interact with various molecules to influence multiple signaling pathways. PEBP-1/RKIP is best known for its role as an inhibitor of MAPK signaling through its interaction with Raf-1, but it also modulates GRK2, NF-κB, and GSK3β. Other reported roles are in cell cycle progression, differentiation, migration, and apoptosis. It also contributes to many physiological functions at the organismal level, including neural development and plasticity, circadian rhythms, immune response, cardiac function, and others. Finally, its expression impacts the development and progression of various diseases, not the least of which is cancer. However, many outstanding questions remain, including how pebp-1/rkip expression is regulated, which downstream signaling pathways are relevant to its effects in the various physiological systems or disease states studied so far, and whether there are additional signaling pathways that are under its regulation.
PEBP-1/RKIP has emerged as a novel class of cell signaling “gatekeeper” that maintains homeostasis by interacting with key signaling effectors and preventing them from throwing the physiological state off balance. Furthermore, PEBP-1/RKIP seems to be a convergence node for various signaling pathways to cross-talk and influence each other. The strategic positioning of PEBP-1/RKIP could be a mechanism by which context-dependent response specificity is achieved.